Abstract
Background: The clinical outcomes of posterior cruciate ligament reconstruction are varied. No previous studies have investigated the effect of graft length on the structural properties of the graft.
Hypothesis: Graft length significantly affects the structural properties of posterior cruciate ligament grafts.
Study Design: Controlled laboratory study.
Methods: Eight Achilles tendon grafts were tested under tensile loads up to 400 N at 3 different lengths: long (75 mm), medium (48 mm), and short (34 mm). These 3 lengths represent midtunnel fixation, inlay fixation, and fixation near the ligament insertions.
Results: Shortening the graft from both long to medium and from medium to short increased the stiffness by approximately 25%. Long and medium grafts displaced significantly more than medium and short grafts, respectively.
Conclusion: The effective length of a graft, which is determined by where it is fixed, should be considered an important variable in posterior cruciate ligament reconstruction.
Keywords
The clinical outcomes of PCL reconstruction have been varied.7,16,20 For example, it has been reported that 60% of patients develop long-term degenerative changes in the patellofemoral and medial compartments after PCL reconstruction. 16 Abnormal kinematics have been thought to contribute to joint degeneration by altering joint contact pressures, leading to an increased rate of cartilage degeneration.2,10 Recently, different reconstruction techniques and graft materials have been proposed to improve the outcomes of PCL reconstruction, including applying an initial tension to the graft,3,8,19 using 2-bundle reconstructions,7,17,25 varying the position of the graft tunnels on the femur and tibia,4,18 and inlaying the bone block of the graft into the posterior tibia.1,21 These techniques have been shown to restore the posterior stability of the knee under relatively low (approximately 100 N) posterior loads applied to the tibia during in vitro studies. The ability of these techniques in restoring normal knee kinematics under in vivo conditions is unknown. However, a recent in vitro study by Gill et al 6 has suggested that despite restoring posterior stability under posterior loads, a single- bundle transtibial tunnel PCL reconstruction does not consistently restore normal knee kinematics under more complex physiological loading conditions.
To restore normal knee kinematics after PCL reconstruction, the graft must reproduce the structural properties of the intact PCL. One variable that directly affects the load-elongation behavior of the graft after reconstruction is the length of the graft between fixation points (effective graft length). Basic structural mechanics theory indicates that a longer effective graft length will result in a less stiff reconstruction than a shorter graft. However, there are no studies in the literature reporting the effect of graft length on the force-elongation behavior. Therefore, this study was designed to investigate the effect of graft length on the force-elongation behavior of an Achilles tendon graft.
Methods
Eight cadaveric Achilles tendons and their bony attachments were used in this study. The tendons were of allograft transplant quality and were prepared in a manner similar to the techniques used clinically. 6 After thawing, the tendon and calcaneal attachment were removed and “tubed” to fashion a graft that would fit an 11-mm tunnel. They were then wrapped in a towel moistened with physiological saline, placed in a resealable plastic bag, and frozen at −20°C for storage. The grafts were thawed at room temperature directly before testing.
The biomechanical testing was performed in a materials testing machine (MTS, Canton, Mass) (Figure 1). To grip the graft, two 7 × 3 × 2-cm custom-made sinusoidal aluminum clamps were used to grip each end of the tendon with 80-grain sandpaper glued to the surface of the clamp. Rubber strips were fixed adjacent to the sandpaper to reduce the amount of deformation of the graft in the lateral direction inside the grip. These strips allowed the tendon to retain its tubular shape within the clamp under high compression forces. Each end of the tendon was secured in the clamp and manually tightened. The graft was hydrated with physiological saline throughout the test.
Achilles tendon graft mounted on materials testing machine.
For each graft, 3 different lengths were studied by sequentially shortening the graft: 75 mm (long), 48 mm (medium), and 34 mm (short). The long, medium, and short grafts represented midtunnel fixation on the femur and tibia in a transtibial tunnel reconstruction technique (78 mm), inlay fixation on the tibia and midtunnel fixation on the femur (48 mm), and fixation near the articular surfaces of the tibia and femur (34 mm), respectively. The long length was tested first, followed by the medium and short lengths, with 30 minutes of graft recovery between each test (wrapped in saline-soaked gauze).
To minimize viscoelastic effects, the graft was elongated between 0 mm and 2 mm at 20 mm/min for 10 cycles before each test. 26 A plot of force versus time during the preconditioning of a typical specimen of medium length is shown in Figure 2. As the number of cycles increased, the peak force decreased with time. This effect decreased with the number of cycles. This trend was observed in the other specimens and other lengths.
Force versus time during the preconditioning of a medium graft. Minimal change was observed after 10 cycles.
The graft was then displaced at a rate of 100 mm/min until a load of 400 N was reached. To ensure that the graft was not slipping in the clamps, the interface of the graft and clamp was monitored visually using ink marks on the surface of the graft. A computer recorded the applied force and the grip-to-grip elongation of the graft during the test. The linear stiffness of each specimen at each length was calculated from the linear region of the force-elongation curve.
A 2-way repeated measures analysis of variance (ANOVA) was used to detect differences in displacement between grafts of different lengths under different levels of loading. A 1-way repeated measures ANOVA was used to detect differences in the linear stiffness of short, medium, and long grafts. Student-Newman-Keuls post hoc testing was used to isolate differences between groups where appropriate. Statistical significance was set at P < .05.
Results
The load-elongation curve for a typical specimen with short, medium, and long lengths is shown in Figure 3. As the force increased, the elongation increased at all lengths. Both a nonlinear toe region and a linear region were observed at all lengths. The longer graft had a larger overall displacement under the same load as compared to the shorter grafts in both the nonlinear and the linear regions. Similar trends were observed in the other specimens.
Elongation versus force curve for a typical specimen. The specimen was tested sequentially from long to short.
The mean elongation versus force plot is shown in Figure 4. The 2-way repeated measures ANOVA indicated that both load and length had statistically significant effects on the elongation of the grafts. Statistically significant interaction effects were detected between load and length. The elongation of the medium-length grafts was statistically greater than that of the short grafts beyond 150 N of force. The long grafts displaced statistically more than the medium grafts beyond 25 N. For example, at 100 N of applied load, the mean elongation of the short grafts was 2.9 ± 0.4 mm (mean ± SD). The elongation of the medium grafts under 100 N of load (3.3 ± 0.4 mm) was not statistically different from that of the short grafts. However, the displacement of the long grafts (4.4 ± 0.6 mm) was statistically greater than the short and medium grafts. At 400 N of applied load, the long graft displaced 8.9 ± 1.2 mm, the medium graft displaced 6.5 ± 0.8 mm, and the short graft displaced 5.4 ± 0.7 mm. All of these differences were statistically significant.
Mean elongation versus force for short, medium, and long grafts (mean ± SE). The plus sign indicates a statistically significant difference between long and medium grafts, and the asterisk denotes a statistically significant difference between medium and short grafts.
The linear stiffness of the short, medium, and long grafts is shown in Figure 5. The stiffness of the short grafts averaged 154 ± 42 N/mm (mean ± SD). The stiffness of the medium grafts (116 ± 37 N/mm) was statistically less than the short grafts. Long grafts had a mean stiffness of 88 ± 36 N/mm, which was statistically less than the medium grafts. Lengthening the grafts from short to medium and from medium to long each resulted in a decrease of stiffness of approximately 25%.
Linear stiffness of short, medium, and long grafts (mean ± SD). With increasing length, the linear stiffness of the graft decreased significantly. *P < .05.
Discussion
Many recent biomechanical studies in the literature have focused on improving the outcomes of PCL reconstruction.3–5,7,17,20 These studies have examined a number of different surgical variables, including the position of the graft tunnels on the tibia and femur, the amount of pretension in the graft, double-bundle reconstruction, and tibial inlay reconstruction. Currently, there are no surgical guidelines indicating where the graft fixation should be placed during PCL reconstruction (eg, at the anterior tibial cortex vs the posterior of the aspect of the tunnel). A previous study from our laboratory using a theoretical model indicated that effective graft length (the length of the graft between fixation points) plays an important role on the structural properties of the PCL graft. 15
The current study used Achilles tendon grafts to quantify the effect of graft length on its force-elongation behavior. Three different effective graft lengths were studied to simulate 3 different graft fixation levels: fixation at the midpoint of the tibial and femoral tunnels in a transtibial tunnel technique (long, ~75 mm), inlay fixation on the tibial side and fixation at the midpoint of the femur (medium, ~48 mm), and fixation at the articular surface of the tibia and femur (short, ~32 mm). Decreasing the length of the graft from long to medium and from medium to short statistically increased the linear stiffness of the graft. To restore the kinematics of the knee after reconstruction of the PCL, the structural properties of the graft being used to replace the PCL must match the properties of the intact PCL. Because the fixation site of the graft in the reconstruction determines the effective length of the graft, these results indicate that the location of the fixation site of the graft plays an important role in the kinematic response of the knee. The optimal fixation site will depend on the graft material and surgical techniques being used to reconstruct the PCL.
Previous biomechanical studies have shown that PCL reconstruction restores the function of the PCL under low levels of posterior tibial loading (approximately 100 N).7,8,17,18,21,25 The current study indicated that at lower levels of loading, the differences in elongation between the different graft lengths were relatively small. For example, at 50 N of applied force, the elongation of the medium graft was 16% more than that of the short graft, and the elongation of the long graft was 27% more than the medium graft. However, with increasing loads, the differences in elongation between the grafts increased dramatically. At 400 N, the percentage increase in elongation was 22% and 36% for the medium and long grafts, respectively. In a recent theoretical study, Li et al 15 demonstrated that applying a pretension to the graft can reproduce the intact ligament tension at low levels of loading, but at higher levels of loading, the graft will transmit less force than the intact ligament because of the nonlinear force-elongation behavior of the graft. Although there are little data on in vivo PCL function, it is assumed that the force in the PCL under physiological loading conditions is considerably larger than the low forces (approximately 100 N) measured during the posterior drawer test. Therefore, using an appropriate graft length to reproduce the structural properties of the PCL across a large range of loading may be an important variable in PCL reconstruction.
Ishibashi et al 11 studied the effect of graft fixation site on the stability of the tibia under anterior drawer loading after ACL reconstruction. They studied 3 different fixation methods on the tibia using a patellar tendon graft: external fixation near the distal end of the tibial tunnel using staples (distal fixation), fixation at the midpoint of the tibial tunnel with an interference screw (central fixation), and fixation near the proximal end of the tibial tunnel using an interference screw (proximal fixation). They measured a progressive decrease in anterior translation with distal, central, and proximal fixation, respectively. These results qualitatively agree well with ours in that a longer graft results in a less stiff reconstruction than a short one.
To our knowledge, no previous studies have quantified the structural properties of the Achilles tendon graft used in PCL reconstruction. Few studies have measured the material properties of the Achilles tendon.13,14,28,29 These studies measured elastic moduli between approximately 375 MPa and 820 MPa. It is difficult to compare our data to these studies because of differences between the experiments. The Achilles tendon graft was constructed by rolling the tendon to form an approximately cylindrical structure, whereas in previous studies the whole Achilles tendon was tested. Also, in our study we measured the displacement of the grip instead of local tissue strain. Studies measuring local strain consistently report higher values of stiffness than studies measuring grip-to-grip displacement. In their study of the tensile properties of the Achilles tendon, Wren et al 29 measured 20% to 30% of the total strain at the bony insertion. We believe that measuring the grip-to-grip displacement is more relevant than local strain measurements in the study of PCL reconstruction because the entire graft is implanted during surgery, rather than just an isolated section of the graft.
Previous investigators have measured the structural properties of the functional bundles of the PCL.9,24 Harner et al 9 measured a linear stiffness of 120 N/mm for the anterolateral bundle of the PCL. If the goal of single- bundle PCL reconstruction is to restore the stiffness of the anterolateral bundle, our data suggest that the graft should be fixed such that the graft length is between the medium and short grafts used in this study. This may be accomplished by using the inlay technique or by fixing the graft used in a transtibial tunnel reconstruction more closely to the articular surface of the joint.
The current study only applied tensile loads of up to 400 N. This level of loading was used because the same grafts were used at 3 different lengths, and we wanted to ensure that the grafts would not be damaged during testing. In this study, the grafts were cyclically preconditioned before each test and allowed to recover for 30 minutes in saline-soaked gauze before the next test. Previous studies in the literature suggest that this protocol is appropriate to minimize the viscoelastic effects of the graft.12,22,23 This experiment used the same amount of recovery time as used in the experiment of Johnson et al. 12 Panjabi et al 22 found no differences in the tensile and viscoelastic properties of the rabbit ACL after repeatedly testing the same ACL with cyclic preconditioning before each test and a recovery period after each test. Furthermore, Provenzano et al 23 have suggested that 10 times the length of the test is a sufficient amount of time for a ligament to recover. In the current experiment, this requirement was easily satisfied.
This study did not investigate the viscoelastic properties of the graft. Graft length may also affect the viscoelastic behavior of the structure. A recent study by Thornton et al 27 has suggested that an abnormal creep response of a graft may contribute to the persistent joint laxity noted after ligament reconstruction. Future studies are needed to quantify the effect of effective graft length on the structural viscoelastic properties of the graft to best match the force produced by the intact PCL.
This study also did not simulate the actual implantation of the graft into the knee and represented the structural properties of the graft shortly after surgery. Some of the variables that may affect the structural behavior of the graft in an actual reconstruction may include differences in bone quality along the length of the tunnel, interactions between the graft and tunnel wall, the type of fixation used, and healing of the graft within the bone tunnel. Future studies may be needed to further quantify the effects of these variables.
In conclusion, the results of this study demonstrate that graft length has an important effect on its structural properties. Therefore, the length of a graft should be treated as an important variable in the reconstruction of the PCL. The optimal length of a graft should be determined by matching the structural properties of the graft to that of the intact PCL. This variable should be carefully considered when performing ligament reconstructions to more closely reproduce the function of the ligament being replaced.